mcpdirectory.dev
9,129 servers · updated 23 Sept

Browser Use MCP server

official from Browser Use Active last commit 17 d
{
  "mcpServers": {
    "browser-use": {
      "command": "uvx",
      "args": [
        "browser-use"
      ],
      "env": {
        "BROWSER_USE_API_KEY": "<browser_use_api_key>",
        "GOOGLE_API_KEY": "<google_api_key>",
        "ANTHROPIC_API_KEY": "sk-ant-..."
      }
    }
  }
}
file .mcp.json (in your project root)install uvx browser-useversion 0.13.5

What it does

Browser Use is an MCP server that allows an AI agent to control a real Chrome browser to complete web tasks like filling forms, extracting data, and navigating websites. It is for developers who want to automate web interactions programmatically or integrate browser automation into AI agents.

  • Fill out web forms and applications automatically with provided information.
  • Extract structured data from web pages and export it in formats like CSV.
  • Navigate websites and complete multi-step tasks like booking or searching.

Tools this server exposes

ToolDoes
browser_controlControl a Chrome browser to open pages, click elements, type text, and interact with web content.
data_extractionExtract and structure data from web pages for export or processing.
form_fillingAutomatically fill in web forms with provided information.

Before you enable it

An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.